When bacteria invade host cells, they may use the host's own cell machinery to infect and persist. Bacteria also use host molecules as signals. These signal exchanges are critical to pathogenesis and to symbiosis. We study Sinorhizobium meliloti, an alpha- proteobacterium, which establishes a beneficial nitrogen fixing symbiosis with host plants including the model legume Medicago truncatula. Molecular study of this bacteria's invasion into its host has yielded basic information and technical approaches applicable to series mammalian diseases such as Brucella. This biological process is also important because it allows certain plant crops such as soybean and pea to grow without exogenous N-fertilizer: the process is thus critical to human and animal nutrition. The interaction of bacterium and host plant is characterized by species specificity and by the complexity of a developmental process in which (1) the host build a root nodule, which is a multi-tissue organ, and (2) the bacteria invade through layers of nodule and occupy target host cells as a long-term intracellular symbiont. We propose to carry out a comprehensive project aimed to identify genes, regulators and signals involved in intermediate stages of symbiosis (infection, invasion, and release into target cells). A particular advantage of this system is that the bacterium and the host can each be grown independently, and can be used for forward genetic screens. This has yielded a bounty of mutants in both the bacterium and in the host, and our project has the special prospect of using plant developmental mutants as an explicit tool for analyzing the bacteria's mechanisms of infection and differentiation within the host tissues. We aim to analyze how the bacteria differentiate during several sequential steps of symbiosis. We will build on some unexpected recent results, including (1) the finding that specific RNA polymerase sigma factors are required for successful symbiosis (2) the observation that some previously known regulators have unexpected functions in multiple bacterial behaviors, and (3) the discovery that one of our plant mutants is defective in producing host peptides that are required for expression of bacterial symbiosis genes. This work will yield insights into how bacteria infect host cells.